In this project I'm going to study the harder styles music in order to justify it's diverse subgenres.
To start the project. There needs to be some grounded evidence that Spotify API is able to distinguish between harder styles and general pop music. So the first steps I took was look at the spotify variables that were unusual or interesting to look at.
Now to load the playlistsThe harder styles music is to some people quite overwhelming. A 'wall of sound' is how this music is often described. So it would be interesting to see the energy associated with harder styles music.
Figure 1 shows that harder styles music does live up to its reputation in this case having a high energy rating associated with fast, loud and entropic music.Energy is dependent on how fast, loud and entropic a track is. So the next interesting thing to discover is how the relationship between loudness and energy is and how the subgenres score in comparison to eachother.
Figure 2 shows energy seems to positively correlate with loudness. Already some slight differentiation between genres is visible. Doubleclick on a subgenre to see it's 'territory' of loudness and energy.But to classify the harder styles on basis of loudness is not always waterproof. For example hardcore, rawstyle and frenchcore are three subgenres that sound very different musically (I might add sound samples in later portfolios to show what I mean, I feel like that is a necessity to get a feel for the differences). That definitely isn't represented in this figure.
Some characteristics of harder styles music can be visually shown using the spotify API. There are some differences and similarities versus mainstream music. Energy and loudness ratings of harder styles were on average higher than top 1000 music. But tempo in BMP was on average pretty simular, which suprised me to be honest.
Between subgenres some characteristics could also be defined. For example loudness of average differs between subgenres aswell. Happy hardcore is less loud than terror, hardstyle is less loud than Frenchcore. Also apres ski and happy hardcore is on average more happy than the other genres.
These characteristics could be rough guidelines to shape a profile of each subgenre. Still some extra work can be done to define them better (and make a table at the end so that you get some overview).
It might also be interesting to delve deeper into specific tracks using the time-tracker. Thus far I completely ignored the structure of the tracks itself. Harder styles music has a pretty generic EDM structure (assuming you might be familiar with it or recognize it) with a buildup towards the drop. But there are some interesting characteristics in those aswell, for example a drop following a drop (I can imagine this is hard to imagine but once I get sound samples you get the idea).
So there are still some interesting aspects to uncover.
TODO: - Continue developing the profiles of each harder style - Might focus more on individual numbers of different genres to try and define them. (Fourier analysis?) - Might look at the artists. - Improve the figures (definitely) - Add table with all genres and characteristics